Vol. 59.] IN GLASSY IGNEOUS ROCKS. 431 



them; except to call attention to the peculiar feathery crystals, 

 especially characterizing those patches of the rock in which a 

 radial growth is inconspicuous. When most perfectly developed, 

 this peculiar structure is but a large variety of the branching rods 

 of felspar of which a good figure is given in pi. xvii, fig. 2, of the 

 7th Ann. Rep. U.S. Geol. Surv. (1885-86). 



In the type under consideration the rods are broader, in com- 

 parison with their length, than those figured ; and, while more curved 

 in outline, exhibit branches which are more stumpy and less con- 

 spicuous (PI. XXVI, fig. 1). Although this mineral is often free 

 from the scattered black microliths,yefcnot uncommonly an example 

 is crossed by regularly-spaced rows of these earliest of crystallizations, 

 frequently arranged with a slight outward convexity, suggesting a 

 forward pushing of foreign substances, which, finally, were perforce 

 engulfed. Often, when most irregular in outline, these crystalliza- 

 tions are composite, breaking up into a granular mosaic as the stage 

 is slowly turned, extinction proceeding in consecutive and regular 

 jerks from grain to grain for the entire length or width of the area 

 under consideration. I regard this as implying that the molecules 

 of the various component grains are orientated in nearly similar 

 directions, each being surrounded by a zone of doubtful polarization in 

 which the change takes place from the direction of orientation of 

 the molecules of one grain to that adopted by the molecules of the 

 adjacent grain. Flow-structure of ten complicates the crystallization 

 of the rock. A comparatively-coarse translucent mineral — no doubt 

 usually tridymite — commonly enters largely into the composition, 

 while in an occasional instance of the ' granophyre-groups,' the 

 black microlith and a confused mass of small spherulites form the 

 rest of the field. The ' feathery growth ' appears locally, and con- 

 stitutes a semispherulitic patch or an entire band in the rock. As 

 in other spherulites, a slight line of discontinuity apparently favours 

 growth. 



As regards the origin of these 4 feathers/ their occurrence in 

 the ; porous patches ' associated with tridymite forcibly suggests a 

 connection with superheated steam. Their shape curiously re- 

 sembles the frosting of glass or pavements ; and their similarity to 

 the branching felspar-rods, figured by Prof. Iddings, suggests a com- 

 munity of origin. We both consider the structure as a result 

 of resistance to growth, and it is discussed at greater length 

 in Prof. Bonney's part of this paper. 



Conditions which favoured Primary Devitrification at 

 Obsidian Cliff. 



The fact that the spherulites formed in far greater proportion in 

 the upper part of the lava-flow which now makes Obsidian Cliff, 

 points to a low pressure with active water-content, an initially- 

 high temperature and rapid fall, as being physical conditions 

 requisite to their production. This disposition of the spherulites, 

 among which the hollow variety is conspicuous, is very marked 

 in looking at the exposure as a whole, especially as the columnar 



2i2 



